MXPA03002033A

MXPA03002033A - A method of controlling an electric motor, a system for controlling an electric motor and an electric motor.

Info

Publication number: MXPA03002033A
Application number: MXPA03002033A
Authority: MX
Inventors: Roberto Andrich
Original assignee: Brasil Compressores Sa
Priority date: 2000-09-08
Filing date: 2001-08-27
Publication date: 2004-12-13
Also published as: HUP0302694A2; EP1317794A2; CN1473390B; US20040032230A1; DE60124121T2; US6922027B2; WO2002021675A3; CN1473390A; HU226334B1; EP1317794B8; AU2001281610A1; WO2002021675A2; ATE343866T1; EP1317794B1; SK2682003A3; JP2004508000A; BR0004062A; JP4778666B2; AR031398A1; ES2275708T3

Abstract

A system of controlling an electric motor, a digital method of controlling an electric motor and an electric motor are described. The system of controlling an electric motor of N phases comprises a microcontroller (10), an A/C converter (30) associated to the microcontroller (30), a set of voltage meters (DN) associated to the A/D converter (30), a set of switches (SW2N) connected to an electric voltage (VBARR) and associated to the microcontroller (30), the microcontroller (30) selectively feeding, by means of at least two switches (SW2N) two phases (FN) of the motor (20) with the voltage (VBARR) during a period of time (TPOS), the movement of the motor inducing the electric voltages (EN), the set of meters (DN) measuring the signals of electric voltages (fN) and comparing these voltages with each other to determine the period of time (TPOS). One describes also the calculation of a parameter, called H (r), used to adjust the instants of commutation according to the constructive type of the motor. The parameter depends upon the rotation of the motor. One also foresees the use of a digital filtering technique, to eliminate the effects of the modulation on the control method. The filtering is characterized by the arithmetic mean of the last k samples of each average electric voltage (fN), these samples being synchronized with the period of modulation.

Description

ELECTRIC MOTOR, METHOD AND SYSTEM TO CONTROL THE SAME.

The present invention relates to a system for controlling an electric motor of a digital control method of an electric motor, particularly of a Brushless CD-type permanent magnet motor, as well as to an electric motor supplied with a digital control system. Description of the Prior Art A Brushless CD-type permanent magnet motor can be operated without any information regarding its position, so it operates in a similar way to an induction motor. However, in order to achieve maximum torsion and efficiency, the phase currents have to be synchronized with the induced voltages. This can be done either by means of sensors physically coupled to the motor, such as for example Hall type, optical sensors, etc., or by observing the induced voltages and / or currents. Coupled sensors have the drawback of adding extra elements to the motor design, considerably increasing the final cost. In addition, the limitations of space and environment to which the engine is subjected may make the use of these types of sensors unfeasible. Therefore, in most cases the best option is the use of voltage and / or current observers. EF: 145610 Examples of control techniques that use voltage and / or current observers can be found in Brazilian Patent No. PI 9904253 of SCH ARZ and associates, in US Patent No. 4, 162, 35 of WRIGHT, US Patent No. 4,169,990 of LERDMAN, US Patent No. 4,743,815 of GEE and associates, US Patent No. 4,912,378 of VUKOSAVIC, US Patent No. 4,928,043 of PLUNKETT, US Patent No. 5,028,852 of DUNFIELD, and US Patent No. 5,420,492 of SOOD, and also in the publications of SATOSHI 1991, SHOUSE 1998, ERTUGRUL 1998 and BOLOGNANI 1999. In the WRIGHT patent, an integration technique is used to determine the switching moment. The non-energized winding is integrated, so the magnetic flux is obtained which is compared with a reference value. When the value of the integral exceeds the reference value, the commutation is made, and the value of the integral returns to its initial value. A disadvantage of this technique is the large amount of hardware that is required for its implementation. The solution is not microprocessed. In the LERDMAN patent, a technique similar to that presented in the WRIGHT patent is used. The non-energized winding is integrated and compared to a reference. A disadvantage of this technique is also the large amount of hardware required. The solution is not microprocessed. In the SWARZ patent, a technique is used that uses an observer of the voltage value constituted by a network, formed by voltage comparators, capacitors and resistors. If this system is balanced, with this observer the phase voltages are taken, thus forming a virtual neutral (zero induced voltage), then the zero is compared with each phase, generating a signal that is 30 ° ahead of the moment of current of the change of position. Subsequently, in order to achieve the correct instants of the change of position (switching), this advance is compensated by a network of capacitors. The solution is microprocessed, but it requires many external components. In the GEE patent, a technique is used to detect the zero of the induced voltage (zero crossing), where a zero is considered as half the voltage value of the CD bar. The voltage in the non-energized winding is compared to zero. The zero is always reached at 30 ° of the switching moment. In this way, a delay is left after detection to determine the switching moment. The cycle formed by the resistors, capacitors and comparators is used in the detection process. There is no modulation in the motor speed control. Rather, an SCR is used that controls the bus voltage. The solution is microprocessed. A disadvantage of this technique is the need to use voltage comparators. Another disadvantage is that in motors having a reduced number of slots in the stator (for example 6-slot rotor + 4-pole, 9-slot rotor + 6-pole), the zero crossing has a flat-region which makes the determination difficult. of the exact moment in which it must occur. In the SATOSHI patent, a zero detection method is always used. In this case, two diodes are used for detection. A diode is connected to each phase. When a phase is not energized, the diode current is monitored. The moment in which the current of the diode is extinguished or the moment in which the current begins to circulate through the diode, represents the zero of the voltage. In the VUKOSAVIC patent, the third harmonic voltage of the motor is achieved by means of the sum of the voltages of the phases. The switching moment is subsequently determined as a function of the phase angle of the third harmonic. Here an advantage is that the signal of the third harmonic does not go through any distortion in the case of voltage modulation in the motor. A disadvantage is the need to access the neutral point of the star connection of the motor. The solution is microprocessed, although another disadvantage is the large amount of external hardware required by the microprocessor.

In the PLUNIETT patent, such as in the WRIGHT and LERDMAN patent, the integration of a non-energized phase voltage is performed. The value of this integration, which represents the magnetic flux, is compared to a reference value called Null Point. This point delimits the voltage signal (zero voltage). For example, if the non-energized phase is on the rise (towards the bus voltage), then the voltage values to the left of the Null Point are considered negative and those to the right are positive. In this case, when the voltage integration process begins, the voltage begins to rise in a negative way until it reaches its maximum negative value at the null point. After the Null Point, it begins to reduce the value of the integral module and the switching moment occurs when the value of the integral reaches zero. The solution is not microprocessed. Here a disadvantage is the large amount of hardware required. In the DUNFIELD patent, high frequency signals are injected into the non-energized phase and the resulting peaks are measured. On the basis of the measured values, the switching moment is determined. In the SOOD patent, a method is used that is very different from those cited above. Here it is not necessary to have the voltages ready in each phase. Only the current circulating through the CD bus as information is used. The motor is initially driven by voltage setting. Subsequently, the switching moment is adjusted according to the shape of the reading current. A microprocessor is used to perform the analysis of the current format. There are also some microprocessors on the market designed to control motors, which have projected peripheries to determine the switching moment. As examples, the ST72141 microcontroller from STMicroelectronics and the TMP88PH47, TMP88PH48, TMP88PH49 microcontrollers from Toshiba can be mentioned. All these microcontrollers use the method to detect the zero, for the determination of the switching moment. In Toshiba microcontrollers, zero is considered half the value of the DC bus voltage, but it needs comparators, resistors and external capacitors to aid in detection. In STMicroelectronics microcontroller, the zero is the reference of the digital circuit. It always needs the presence of modulation in the phases, because the reading of the voltage of the non-energized phase must be made when all the connections of the inverter are open. Only three 3 resistors and 3 capacitors are required to aid in detection.

BRIEF DESCRIPTION OF THE OBJECTIVES OF THE INVENTION In the technique proposed in the present invention, the voltages in the three phases of the motor are sampled, treated in mathematical form, added to a parameter proportional to the speed of the motor and depend basically on the form that constitutes this engine and compare each other. The result of this comparison determines the moment of switching of the phases. The switching moment can be advanced or delayed by only changing the parameter by means of software. All the part of the control and perception of the position, is carried out only by means of a Digital Signal Processor or an equivalent circuit, here defined as the assembly of the microcontroller associated with an analog to digital (A / D) converter. The control system and method of the present invention have the objective of eliminating two analogous circuits to determine the position of the rotor and to seek the action of the motor with the correct angle between the current and the voltage imposed on the windings of this motor, allowing the control of this angle by means of a parameter inserted in the software. Another objective of the present invention is to allow the detection of the position of the rotor for higher powers, even in situations in which demagnetization ends after 30 ° electrical from the moment of the last switching, that is, after the moment of zero crossing of the voltage in the non-driven phase. The method and system also have the objective of covering a wide range of rotation, allowing a total torsion from 2% of the maximum rotation (Below this value the voltages at the entrance of the meters are very low). The system and method have the additional objective of using only a digital signal processor and three resistance dividers with first-range CR filters to read the voltages in the motor phases, without necessarily requiring the presence of modulation, for example, MAP (pulsation width modulation) of the voltage in the f ses. Another objective of the proposed method and system is to accept trapezoidal shapes of induced voltages with a level lower than 120 °, which are found when different forms of motor construction are used. Another objective of the proposed method and system is to accept the MAP modulation in the phases of the engine with a cyclic proportion of 100%. An additional objective of the proposed method and system is to operate with the control technique through both the imposition of voltage and the imposition of current on the motor windings. One of the objects of the present invention is achieved by means of a method for controlling the N-phase permanent magnet electric motor, which comprises a microcontroller with a group of voltage meters associated with the microcontroller, a group of connections connected to a electrical voltage and associated to the microcontroller, selectively driving the microcontroller at least a pair of connections, applying a voltage to at least two phases of the motor, the method comprising the steps of reading the microcontroller by means of a group of meters, corresponding the signals of the electric voltages to the supply voltages in the phases of the motor, and comparing the microcontroller the voltages of the phases among themselves and with pre-established parameters and driving at least a new pair of connections, as soon as the relations are satisfied pre-established by means of voltages. Another objective of the present invention is to achieve by means of a control system of a permanent magnet electric motor of N phases comprising a microcontroller, a group of connections connected to an electrical voltage and associated with the microcontroller, operating selectively the microcontroller at least a pair of connections, applying a voltage to at least two phases of the motor, the system comprising a group of voltage meters associated with the microcontroller, the group of meters being connected to the power input of the motor phases . These microcontrollers comprise stored in their memory, pre-established relationships between the voltages, and have the ability to compare the measured value by means of the meters with the pre-established relationships and drive at least an additional pair of connections depending on the voltages measured by the group. A further objective of the present invention is achieved by means of an electric motor of permanent magnet of phases N comprising a system with the ability to synchronize the phase currents with the induced voltages, which includes a microcontroller, a group of connected connections to an electrical voltage and associated with the microcontroller, selectively driving the microcontroller at least one pair of connections, applying a voltage to at least one group of voltage meters associated with the microcontroller, a group of meters being connected to the inputs of the microcontroller. power supply to the motor phases. The microcontroller includes, stored in its memory, pre-established relationships between the voltages and has the ability to compare the value measured by the meters with the pre-established relationships and drive at least an additional pair of connections depending on the voltages measured by the group . A further objective of the present invention, is achieved by means of a method to control the position of the rotor of an electric motor of permanent magnet of phases N, the motor being fed by a group of connections selectively switched by a microcontroller, the method comprising the use of a parameter called H (r) provide the constructive factors of the motor, provide the motor rotation and provide the scale factor of the voltage meters that will be used as an adjustment factor in the process to compare the phases, to determine the moment of switching with the maximum combination of motor connections. A further objective of the present invention is achieved by means of a method for controlling an electric motor of permanent magnet of phases N, the motor being fed by a group of connections selectively switched by a microcontroller, the method comprising the use of a digital filtering technique to eliminate from the rotor phases the distortions caused by the voltage modulation, for example of the MJVP type. The filtering technique consists of the arithmetic average of the last samples k of the measured voltages, and they are synchronized with the modulation frequency. The sampling range is equal to a multiple of the integer k of the modulation frequency. Brave Description of the Drawings The present invention will be described in more detail below, with reference to a modality represented in the drawings.

Figure la represents a block diagram of a system for driving a three-phase, 4-pole, permanent-magnet type CD Brushless motor with trapezoidal voltages of 120 electrical level and Fig. Ib the respective time diagram; Figure 2 represents an overlap of the induced voltages (EN) per phase, with trapezoidal shape and level of 120 electrical degrees, and the voltage at the common point (VCOMMON) of a permanent magnet motor type CD Brushless; Figure 3 represents an overlap of voltages VN and the voltage at the common point (Vcommon) of the motor indicated in Figure 1 and Figure 7 for the case where the induced voltages (EN) are trapezoidal with a level of 120 electrical degrees. Figure 4 represents an overlap of induced voltages (EN) per phase, with trapezoidal shape and a level less than 120 electrical degrees, and the voltage at the common point (VCOMMON) of a Brushless CD-type motor, this figure also identifies obtaining the parameter H (r); Figure 5 represents a voltage overlap VN and the voltage at the common point (VCOMMON) indicated in figure 1 and the figure in figure 7 for the case where the induced voltages (VN) are trapezoidal with a lower level at 120 electrical degrees. Figure 6 represents one. command signal of the SW2N connections for each position, the voltages at the motor phase inputs (FN) and indicates the switching instants from position 2 to position 3 and from position 3 to position; Figure 7 represents the system used in the present invention, formed by a rectification unit (40), a rectification filter (50), a group of connections SW2N connected between a potential VBARR and the GND to ground, an electric motor of Brushless CD type permanent magnet (20), a group of voltage meters DN, a digital signal processor (10 + (30); Figure 8 represents the command signal of the SW2N connections, the voltage in one of the phases of the motor FN, the voltage to be sampled fN, the voltage VN in the winding of the corresponding motor and the result fNMEDIO (AVERAGEfN) of the mathematical treatment of the samples, when a MAP action is used, also indicating the sampling moments of the voltage fN , according to the system of the present invention: Figure 9, represents in detail the sampling moments of the voltage fN in one of the phases of the engine, as well as the result fNMEDIO of mathematical treatment of the samples d e fN assembly, for a system in which MAP modulation is applied to the voltage in the motor phases. Figure 10a represents the experimental waveforms of a motor with a stator of 6 slots of concentrated windings - three-phase, with 4 poles, as well as figure 10b an extension showing the sampling instant of phase A and the average obtained by the mathematical treatment of the samples. Detailed Description of the Figures Figure 1 (a) shows the basic configuration of an inverter and (b) the ideal waveforms that exist when driving a 4-pole motor of three phases of permanent magnet type CD Brushless, trapezoidal wave . In normal operation, the control analyzes the input of the voltage and / or current observer and operates the connections SW1, ... S 6 in the sequence indicated in figure 1 according to the detected position. In the case of an overcurrent, indicated by the current observer, all connections are opened to protect the system. With reference to Figure 7, the control system of the present invention is carried out entirely by a microcontroller (10), an A / D converter supplied with at least 3 inputs to read the voltages (fN) in the meters (Dn). ) which corresponds to the voltages in the phases (FN). Obviously, the microcontroller (10) can be replaced by an equivalent device having the same characteristics as a microcontroller associated with peripheries and others of a digital signal processor.

Figure 2 shows the ideal overlapping waveforms induced by a Brushless CD-type, three-phase, 4-pole permanent magnet motor with trapezoidal voltage and 120 electrical degrees. The level is defined as the angle in electrical degrees, in which the induced voltages remain in a higher value (positive level) or lower (negative level) and approximately constant. In this figure you can see the following relationships between phase voltages for each stage (position) of 60 electric degrees. Position 1 < = > EA > EC > EB Position 2 or EA > EB = Ec Position 3 or EB > EA > Ec Position 4 or EB > Ec > EA Position 5 < = > Ec = EB > EA Position 6 O Ec > EA = EB Table 1 - Relations between voltages induced in the motor In this way, it can be seen that each position has a well-defined relationship between the voltages induced in the motor phases. For example, in the case where the real position is in position 1, position 2 must be started when the voltage induced in phase C (Ec) is equal to the voltage induced in phase B (EB) (see figure 2) and this is less than the voltage induced in phase A (EA). In the same way, position 3 must start when the voltage induced in phase B (EB) is equal to the voltage induced in phase A (EA) and this is greater than the voltage induced in phase C (Ec). With reference to figure 3 and figure 7, it can be observed, with respect to the ground connection of a circuit (GND), the voltages VA, VB, and Ve with the common point of the motor VCOMMON: (1) VA = EA + VCOMMON (2) VB = EB + VCOMMON (3) Ve = EC + VCOMMON The voltage at the common point of the motor (VCOMMON) for the case of the trapezoidal voltage with a level of 120 degrees, is half the value of the bus voltage. (4) VCOMMON = VBARR / 2 Therefore, the voltages VA, VB and Ve are connected symmetrically between the bus voltage (VBARR) and the ground connection (GND). If the speed variation of the motor is made by directly varying the bus voltage (VBARR), that is, without MAP modulation, and if the motor runs in vacuum, these voltages will have the positive level value equal to the bus voltage (VBARR) and the value of the negative level equal to the ground connection value (GND) as illustrated in figure 3. It can be seen in this figure, that the relations between the voltages VA, VB and Ve are equal to the relations between the induced voltages EA, EB, Ec indicated in table 1. Subsequently, the following table can be written. Position 1 or VA = Ve > VB Position 2 or VA > VB = Go Position 3 VB > VA > Go Position 4 or VB > See > VA Position 5 or Ve = VB > VA Position 6 = > See > VA = VB Table 2 - Relationship between voltages VN in the motor phases For a real system, the obtaining of induced voltages with a level of 120 degrees greatly restricts the design and construction of the motor. Therefore, with reference to figure 4, a generic voltage waveform (level below 120 degrees) is considered. In this case, the relationships between the voltages induced for each position, indicated in Table 1, continue as true. However, it should be noted that for the case of a level of 120 degrees, at the time of switching, characterized by the equality of two phases, this equality occurs either with phases at the maximum (positive level) or with minimum value (level negative). On the other hand, in the case of a level below 120 degrees, the equality between two phases (instant of commutation) always occurs at a voltage difference of 2H from the positive or negative level. In figure 5, the voltages of VA, VB and Ve can be observed for the case of a level lower than 120 degrees. The relationships in Table 2 continue as true in this situation. It should be noted that the voltage at the common motor point ((VCOMMON) does not remain fixed at half the bus voltage (VBARR / 2) for any longer time.The reduced level causes an amplitude fluctuation H around this value (VBARR / 2) This distortion in the voltage at the common point of the motor (VCOMOMON), causes the shape of the waves (VN) to be different from the shape of the induced voltages (EN). In this case, the equality between two of the voltages N (instant of communication) always occurs in a difference H of the positive or negative level and not 2H, in the case of induced voltages IN. For the performance of the motor, it is not possible to have simultaneous direct access to the voltages induced by N EN. Therefore, it is not possible to directly use the relationships in Table 1. Furthermore, in order to read these voltages, it may be necessary to have access to the common point of the motor, which makes the circuit of perception and also the project of the expensive motor. The VN voltages that are referenced for the ground connection (GND) can not be accessed simultaneously due to the inductances (LN) and resistors (RN) (see figure 7) of the motor winding. Therefore, the relationships in Table 2 can not be applied directly either. The point of perception used in the present invention are the power inputs of the motor phases (FN) (see figure 7). In order to understand the method of perception using these techniques, figure 6 should be observed. This figure illustrates the waveforms in the FN inputs obtained in the case where the level of the induced voltage of the motor is less than 120 degrees .. Here it is considered that the motor runs in vacuum and without modulation of the bus voltage (VBARR). As an example, when analyzing the switching from position 2 to position 3: when the motor is being driven in position 2, the connection SW1 that connects the input FA to the VBARR bus voltage, the SW6 connection connects the input Fe to GND grounding. The FB entry opens. Therefore, there is no current flow in the resistor (RB) and the inductance (LB) of this winding, and the voltage VB is the voltage value in this input. When adding, in position 2, you get: FA = VBARR FB = VB Fe = 0 (GND) The switching instant or position 3 must occur when the voltage VB is equal to the voltage VA. However, observing Figure 6, it can be seen that at the time of commutation, VA is equal to VBARR-H (considering the engine virtually running in vacuum). In this way, the following relation can be written to be satisfied at the time of switching from position 2 to position 3: Switching 2? 3: FB = FA - H >; Fe Continue in position 3: FA = VA FB = VBARR Fe = O (GND) The instant of commutation to position 4 must occur when the voltage VA is equal to the voltage Ve. Looking at figure 6, it can be seen that in the instant of switching Go, GND + H is valid (considering the motor running virtually empty). Therefore, the following relationship can be written which will be satisfied at the time of switching from position 3 to position 4: Switching 3? 4 FB > Fe + H = FA Extending the same reasoning for the other commutations, the following table is reached: Switching 6? 1 FA = Fe - H > FB Switching 1? 2 FA > FB + H > Faith Switching 2? 3 < = FB = FA - H > Faith Switching 3? 4 FB > Fe + H = FA Switching 4? 5 Fe = FB - H > FA Switching 5? 6 Fe > FA + H = FB Table 3 - Relations between the FN voltages at the switching instants Comparing tables 2 and 3, it can be seen, in the form of differences, the replacement of the voltages VN by the voltages FN the inclusion of the parameter H As the rotation of the motor varies, the amplitude of its induced voltages varies proportionally and as a result the variation of the parameter H. Therefore, once the parameter H is obtained for a rotation ro (H (ro)), can obtain its value for a rotation H by the expression: (5) H (ro) - H (ro) Obtaining the parameter H for a motor, can be done in a very simple way: a) impose a rotation r to the motor , while all SW2N connections are turned off; b) read the induced voltages IN; this is possible in this situation, since there will be no current through the resistances (RN) and inductances (LN) of the windings; c) read the parameter H (ro) as being half the difference between the peak value of the induced voltages (EP) and the voltage E * which corresponds to the voltage at which the voltage module reduced in two phases are equal between yes (see figure); (6) H (ro) = (EP - E *) 12 d) use expression (5) to obtain H (r) for a rotation of r anyone. Until now the engine has always been considered running under vacuum. In the case of normal operation of the motor, the presence of current in the windings originates the voltages induced IN and consequently the VTST voltages pass through a reduction of amplitude in the same rotation r. In this way, the maximum amplitudes of the VN voltages are lower than the bus voltage (VBARR) and their minimum amplitudes are greater than the ground connection (GND). Therefore, the value of H (r) must be increased to compensate for this current variation. This increase must be proportional to the value of the current. If you do not want to use the current value in the control algorithm, multiplication can be added by a constant factor ki > 1, during the obtaining of the parameter (H (r)) in the step d described above to compensate the current variations. You get: (6) H (r) = (r / ro) .H (r) .Ki In this way, with the engine running under vacuum, you always have the switching instant a bit advanced cor-respect to the correct moment . As the motor current increases, the switching time is delayed. The value of Ki can be adjusted experimentally to ensure the proper functioning of the motor in the maximum current condition. As an initial suggestion, the value Ki = 1.3 can be used. If Ki is set to a value lower than 1, there is a reduction of the value H (r) and consequently there will be a delay at the time of switching. Therefore, in addition to being used to compensate for the increase in current, this constant can be used to allow forward and delay the switching time. In order to perceive the algorithm, it is necessary to take into account the scale factor of the acquisition system indicated in figure 7. The DN meters have a scale factor KD, determined by: (7) KD = Rl / (Rl + R2) In this way, the voltages fN are connected to the input of the A / D converter (30) and are determined by (8) fN = KD.F The parameter H (r) graduated by the KD. Adding the scale factor KD to table 3 and reing H with H (r), the following table is reached that will be used in the control algorithm: Switching 6 ~ 1 < = > fa = fe - KD.H (R) > fb Switching 1? 2 < = fa > fb + KD.H (r) = faith Switching 2? 3 fb > fa - KD. H (r) > fc Switching 3? 4 fb > fc + K () = fa Switching 4 - »5 < = > fe = fb - KD.H (r) > fa Switching 5? 6 o Fe > fa + KD.H = fb Table 4 - Relationships that will be tested in microcontroller to perceive the position. In the final application, the variation of the rotation of the motor can be obtained by directly varying the voltage of VBARR (see figure 4), or by modulating the voltage applied to the motor by means of the connections SWl, SW2, ... SW6. For example, in the case of the modulation carried out by means of connections SWl, SW3 and SW5, the voltage in phase A (and also in the remaining ones) will have the form indicated in figure 8. The filter RlC present in DN (see figure 4), smoothes the voltage variations in this phase. In this way, a waveform similar to the one shown in figure 6 is obtained. Here, in order to make the comparisons between the voltage observed in each phase and according to table 4, it is necessary to filter the modulation completely in fa, fb, and faith. For this purpose, the average of the voltages in each modulation period (T) is calculated. The sampling frequency used to measure the voltage fN must be synchronized with the modulation frequency (F). In each period of modulation, always show the values k, separated equally, in each phase, that is, the sampling frequency is determined by: (9) FS = k. F The sum of these values k divided by the number of samples k, represents the approximate average value of the voltage over the period of modulation T. This operation is performed in each sampling of the voltage fN, allowing to obtain the value of fNMEDIO (AVERAGEN ), each voltage sampling cycle fN, that is, at times of frequency k as high as the MAP modulation period of the voltage in the motor phases. In order to avoid the need to wait for a complete period T to obtain the average value available, the following procedure is adopted: in each sampling period TS = 1FS, the sampled value is added to the previous samples k-1 and it is divided the result between k. Speaking in a simpler form, the proposed technique for determining the value of fNMEDIO consists of the arithmetic average of the last samples k. This technique provides an optimal resolution in determining the correct switching moment of the motor, even for relatively low connection MAP frequencies.

As an example, Figure 9 represents the extension of the detail indicated in Figure 8. In this example, 5 samples are considered per period of modulation (k = 5). After the fifth sample of the modulation period n, the average value is: (11) faMEDIO 5 (n) = fa5 (n) + fa4 (n) + fa3 (n) + fa2 (n) + fal (n) 5 Entering the modulation period n + 1, the maximum value of the faMEDIO will be: (12) faMEDIO 1 (n + 1) = fal (n + l) + fa5 (n) + fa4 (n) + fa3 (n) + fa2 (n) 5 and then: (12) faMEDIO 2 (n + 1) = fa2 (n + l) + fal (n + l) + fa5 (n) + fa (n) + fa3 (n) 5 Thus, in each TS sampling period, the available value is available. It is important to note that the cutoff frequency of the filter R1C must be less than half the sampling frequency (Fe <Fs / 2), thus respecting the Nyquist criterion. As a good estimate you can use ¼ less than the sampling frequency. For the sampling frequency you can use, for example, 4 * F or more, in order to have a good average. Therefore, as a suggestion: (14) Fs < = 4.F (15) Fc < = 2.Fs In figure 8, the dotted line indicates the average value of faMEDIO obtained with the proposed technique. It should be noted that the faMEDIO form in Figure 6 is the same as FA in Figure 6, when there is no MAP modulation of the voltage in the motor phase. Later it is found that, with the proposed technique, the effect of the MAP modulation on the voltages of the phases is virtually filtered in its entirety, without presenting a significant disadvantage between the real value of the average and the calculated values, thus allowing to use the ratios indicated in table 4, where the values are subsequently averaged (see table 5) to determine the increase in motor switching. Switching 6? 1 faMEDIO > FcMEDIO - KD. H (R) > fbMEDIO Switching 1 - 2 or faMEDIO > fbMEDIO + KD. H (r) = fCMEDIO Switching 2? 3 < = fbMEDIO > faMEDIO - KD.H (r) > fCMEDIO Switching 3? 4 fbMEDIO > fcMEDIO + K (r) > FaMEDIO Switching 4? 5 FACMEDIO > fbMEDIO - KD. H (r) > FaMEDIO Switching 5? 6 = > fcMEDIO > faMEDIO + KD.H = fbMEDIO Table 5 - Table of final comparisons to be used in the algorithm A strong point in the use of the proposed filtering technique to observe the voltages induced in the phases of the engine, is that this process allows to monitor the position of the rotor times k in each MAP modulation period of the applied voltage in the motor phases. Therefore, even in the case of using a low modulation frequency (Ex. LKHz or less) a good resolution is still achieved during the perception of the rotor position and the appropriate switching moment. As an example, figure 10A should be considered. This figure shows the experimental results achieved in the control of a 6-slot permanent magnet motor, 4-pole Brushless CD type, being F = 1.2KHZ, Fe = 3.3KHz, Fs = 16 ^ F = 19.2KHz. The rotation is 1500rpm. In Figure 10b, we have the extension of 2 MAP cycles, and we also indicate the sampling instants of F a. As previously mentioned, although the frequency of modulation is slow, the mathematical fixation technique proposed here allows to monitor the change of position 16 times in each period T. In this way, an excellent resolution is obtained in the perception of position. It is important to note that the amplitudes of the signals shown in Fig. 1a are at different scales, serving only as a qualitative reference. With reference to Figure 10, the peak voltage observed in phase A after switching 5 - »and the absence of voltage after switching 2 - > 3, characterizes the period of demagnetization of this phase. The demagnetization is the period required for the emergence of the extinction of the current circulating in vff vff vff vvvv a certain phase, after the end of the period of application of the current in this phase, that is, after switching from position x to the next position x + 1. Demagnetization causes a distortion in the waveform of the voltage observed at the motor terminals. In order to prevent this distortion from affecting the perception of the position, there is a certain period of delay TD after each switching, where the switching between the voltages FN is not carried out. An advantage of the method of the present invention compared to the traditional method for detecting the zero of induced voltages is that in the method for detecting zero, the demagnetization period can not exceed 30 electrical degrees, since in this case no you can detect the zero that arises exactly after 30 degrees. In the method of the present invention, demagnetization can exceed 30 degrees without causing any problem in all perception, and for this purpose it is sufficient to conveniently adjust the delay period TD. It is important to note that both the R1C filter and the mathematical calculation process of the average result in a delay in the result. However, this delay, even if it is short and has little influence, can be compensated by adjusting only the parameter H (r). If the PAM modulation is 100%, the VBARR voltage will be continuous and it will not be necessary to carry out the sampling; in this particular case, it is possible to carry out the perception of position by simply observing the voltages in the phases and subsequently comparing these values according to table 4. The sampling range in situations in which sampling is less than 100% , it must be adequate to guarantee a good resolution in the maximum rotation of the motor. In this rotation, the electric frequency of the motor will be determined by: (16) FEL = n. p = 4500 2 = 150Hz where: n = motor rotation in rpm; and p = number of pairs of poles In each electrical period there are 6 different positions, one will be a minimum period per position for this engine: (17) MINIMUMTPOS = 1 = 1.1 lms 6. 150 Hz.

Therefore, a good value to be used to maintain a good resolution at maximum rotation is 10 samples per position. Subsequently the following value for the sampling frequency is achieved: (18) Fs = 10 = 9.09kHz MINIMUMTPOS Obviously, the object of the present invention is applicable to motors of construction similar to that of permanent magnet motors, and having any number of poles and phases, being necessary only to unify them with the particular application. Having described a preferred embodiment, it should be understood that the advancement of the present invention encompasses other possible variations, being limited only by the content of the appended claims, which include the possible equivalents.

It is noted that in relation to this date, the best method known by the applicant to carry out the aforementioned invention, is that which is clear from the present description of the invention.

Claims

Having described the invention as above, the content of the following claims is claimed as property: 1. A method for controlling an electric motor with n phases n (FN), comprises: a microcontroller; - a group of voltage meters (DN) associated with the microcontroller; - a group of connections (SW2N) connected to an electrical voltage (VBARR), and associated with the microcontroller; - selectively activating the microcontroller at least a pair of connections (SW2N), applying a voltage (VBARR) for at least two phases (FN) of the engine, the method being characterized in that it comprises the following stages: the microcontroller reads, by means of from the group of meters (DN), the signals of the electric voltages (fN) that correspond to the supply voltages in the phases (FN) of the motor, and the microcontroller compares the corresponding voltages (fN) with the voltages in the phases ( FN) with pre-established parameters and activates the connections (SW2N) as soon as the pre-established parameters are achieved by the voltages (fN).
2. - A method according to claim 1, characterized in that before the stage of reading the voltages (fN), the values are converted by means of an A / D converter.
3. - A method according to claim 2, characterized in that before the stage of comparing the voltages (fN), an average value (fNMEDIO) is obtained by means of the values (fN), obtained by sampling period (Ts), distributed evenly within the period of time (T).
4. - A method according to claim 3, characterized in that the comparison step includes: comparing the average (fNMEDIO) of the voltage values (fN) shown on each meter (DN) with the average of the voltage values ( FNMEDIO) shown in each of the other meters (DN), subtracting the average value (fNMEDIO) of one of the phases that are being added from a parameter (H (r)) proportional to the motor rotation.
5. A method according to any of claims 1 to 4, characterized in that: - the connections (SW2N) comprise connections (SW1) to (SW6) and the meter (DN) comprises meters (DA) to (DC), and the stage of comparison of the voltages (fN) corresponding to the voltages in the phases ¡FN); - a first combination of connections (SWl) and (SW4) is activated when the average value of the voltage (faMEDIO) in the meter (DA) is greater than or equal to the average value of the voltage (fcMEDIO) in the meter (DC) subtracted from the parameter (h (r)), and the average value of the voltage (fcMEDIO) in the meter (DC) subtracted from the parameter (h (r)), is greater than the average value of the voltage (fbMEDIO) in the meter (DB); - a second combination of connections (SW1) and (SW6) is activated when the average value of the voltage (faMEDIO) in the meter (DA) is greater than the average value of the voltage (fbMEDIO) in the meter (DB) added to the parameter ( h (r)), and the average value of the voltage (fbMEDIO) in the meter (DB) added to the parameter (h (r)) is greater than or equal to the average value of the voltage (fcMEDIO) in the meter (De); a third combination of connections (SW3) and (SW6) is activated when the average value of the voltage * (fbMEDIO) in the meter (DB) is greater than or equal to the average value of the voltage (faMEDIO) in the meter (DB) subtracted from the parameter (h (r)), and the average value of the voltage (faMEDIO) in the meter (DA) subtracted from the parameter (h (r)) is greater than the average value of the voltage (fcMEDIO) in the meter (De); - a fourth combination of connections (SW2) and (SW3) is activated when the average value of the voltage (fbMEDIO) in the meter (DB) is greater than the average value of the voltage (fcMEDIO) in the meter (De) added to the parameter ( h (r)), and the average value of the voltage (fcMEDIO) in the meter (De) added to the parameter (h (r)) is greater than or equal to the average value of the voltage (faMEDIO) in the meter (DA); - a fifth combination of connections (SW2) and (SW5) is activated when the average value of the voltage (fcMEDIO) in the meter (De) is greater than or equal to the average value of the voltage (fbMEDIO) in the meter (DB) subtracted from the parameter (h (r)), and the average value of the voltage (fbMEDIO) in the meter (DB) subtracted from the parameter ((r)) is greater than the average value of the voltage (faMEDIO) in the meter (DA); - a sixth combination of connections (SW4) and (S5) is activated when the average value of the voltage (fcMEDIO) in the meter (De) is greater than the average value of the voltage (faMEDIO) in the meter (DA) added to the parameter (h (r)), and the average value of the voltage (faMEDIO) in the meter (DA) added to the parameter (h (r)) is greater than or equal to the average value of the voltage (fbMEDIO) in the meter (DB);
6. - A method according to claim 5, characterized in that when the combinations of the connections (SW1) to (S6) are activated, the parameter (h (r)) is the result of the multiplication of the parameter (H ( r)) by a scale factor (KD).
7. - A method according to claim 6, characterized in that the step of comparing the average (fNMEDIO) of the values of voltages shown in each meter (DN) with the average of the values of voltages (fNMEDIO) shown in each one of the other meters (DN) is carried out in each sampling period of the voltages (fN) in the meters (DN).
8. - A method according to claim 7, characterized in that the sampling of the voltages (fN) in the meters (DN) corresponding to the voltages in the motor phases (FN) is carried out the times k within the time period T.
9. - A method according to claim 8, characterized in that the period of times T is equal to the MAP modulation period in the motor phases when this MAP modulation is applied.
10. - A method according to claim 9, characterized in that the average value (fNMEDIO) of the voltage shown in each of the meters (DN) of the voltage of the motor phases, is obtained by calculating the arithmetic average of the last samples k.
11. - A method according to claim 10, characterized in that the value (H (ro) is the result of half the difference between the maximum value of an induced voltage (Ep) observed in a phase, in a given rotation (ro) of the motor, and a voltage (E *) observed in two phases, when these values of induced voltage (EN) in the two phases are equal to each other
12. - A method according to claim 11, characterized in that the parameter (H (r)) that will be added or subtracted from the average value (fNMEDIO) of the voltage shown in the voltage meter (DN) of one of the phases is equal to the ratio between the rotation (r) and the rotation ( ro) multiplied by the value (H (ro)) measured in the rotation (ro) and multiplied by the adjustment constant (ki)
13.- A method according to claim 12, characterized in that the use of the parameter (H) (r) is proportional to the rotation (r) to offset the moment of action of the new combination of connections (SW2N).
14. A system for controlling an electric motor of phases N comprising: a microcontroller, a group of connections (SW2N) connected to an electrical voltage (VBARR) and associated with the microcontroller; - Selectively activating the microcontroller at least one pair of connections (SW2N), applying a voltage (VBARR) for at least two phases (FN) of the motor, the system being characterized in that: it comprises a group of voltage meters (DN) associated with the microcontroller, the group of meters (DN) being connected to the power inputs of the phases (FN) of the motor; the microcontroller comprises, stored in the memory, pre-established relationships between the voltages (fN), and have the ability to compare the value measured by the meters (DN) by the pre-established voltages (fN¡ measured by the group). (DN)
15. - A system according to claim 14, characterized in that an A / D converter is associated with the microcontroller to convert the signals read by the group of meters (DN). claim 14 or 15, characterized in that the group of meters (DN) comprises a resistor divider associated with a capacitor., to form an RC filter that has a scale factor (KD) of relationship between voltage (fN) and voltage in phase (FN). 17. - A system in accordance with the rei indication 16, characterized in that a microcontroller reads the voltage value (fN) by means of a sampling frequency (Ts) ve times k as high as the modulation frequency MAP. 18. - An electric motor of phases N, characterized by comprising a system with the ability to synchronize phase currents with induced voltages, which includes a microcontroller, a group of connections (SW2N) connected to a parenthesis (VBARR) and associated with the microcontroller, - selectively driving the microcontroller at least a pair of connections (SW2N), applying a voltage (VBARR) to at least two phases (FN) of the motor, comprising a group of voltage (DN) meters associated with the microcontroller, the group of meters (DN) being connected to the power inputs of the phases (FN) of the motorcycle. - comprising the microcontroller, stored in its memory, pre-established relationships between the voltages (fN) and having the ability to compare the value measured by the meters (DN) with the pre-established relationships and drive at least one additional pair of connections (SW2N) according to the ratio of the voltages (fN) measured by the group (DN).